A New Constraint Handling Approach Based on Enhanced Quadratic Approximation: Tested on Optimal Design of Mechanical Systems and Truss Structures

Abstract

Optimization refers to the process of identifying the optimal state of a system while ensuring all constraints and requirements are met. In engineering problems, the feasibility of solutions is typically assured by imposing relevant constraints. Since these constraints have different properties, utilizing more systematic and logical methods to handle them has the potential to enhance the search performance of the optimization algorithms. According to this fact, in the current study, a new constraint handling mechanism based on combining the fly-back method, weighted average concept and quadratic approximation approach is developed. The proposed mechanism is then merged with three distinct well-established metaheuristic optimization methods. The effectiveness of the enhanced techniques is evaluated through comparative analysis in solving various mathematical and engineering optimization problems subjected to different constraints. Furthermore, non-parametric statistical tests are conducted to compare the quality of the obtained results. The results show that the developed approach can considerably improve the performance of the search algorithms with regards to accuracy, stability, and computational cost.